Parrondo's Paradox

Two LOSING games, when combined, create a WINNING strategy! Discovered by Spanish physicist Juan Parrondo in 1996, this paradox reveals how alternating between two bad choices can somehow produce a good outcome.

The Two Games

Game A LOSING

Simple biased coin flip:
Win with probability 0.495 (49.5%)
Lose with probability 0.505 (50.5%)

Slightly worse than fair — you slowly lose money.

Expected value per round -$0.01

Game B LOSING

Capital-dependent coin flip:
If capital divisible by 3: Win 9.5% (terrible!)
Otherwise: Win 74.5% (great!)

Seems good, but the bad coin hits at the worst times.

Expected value per round -$0.01

A + B Combined WINNING!

Alternating strategy (AABB...):
Play 2 rounds of A, then 2 rounds of B, repeat.

Game A disrupts the "divisible by 3" pattern that makes B lose!

Expected value per round +$0.015

Interactive Simulation

Round: 0 | Strategy: None

Game A Only

Game B Only

AABB Combined

Random A/B

Why Does This Work?

The Key Insight: Games A and B are NOT independent when combined!

Game B loses because you tend to land on "divisible by 3" at bad times. But Game A (the slightly unfair coin) disrupts this pattern.

When you alternate, Game A shifts your capital away from multiples of 3, so when you play Game B, you're more likely to use the GOOD coin (74.5% win rate)!

This is a Brownian ratchet: Random fluctuations from A, combined with B's asymmetric rules, create directed motion — like a ratchet that only turns one way despite random forces.

Real-World Applications

🧬 Cancer Treatment

Alternating between high-dose and low-dose chemotherapy can be more effective than either approach alone (2025 research by Guan et al.)

🦠 Biology

Slime molds use similar strategies — alternating between growth patterns that are individually suboptimal but together optimize survival.

📈 Finance

Portfolio rebalancing between volatile assets can generate returns even when individual assets are expected to decline.

⚡ Physics

Brownian ratchets and molecular motors use thermal noise plus asymmetric potentials to create directed motion.

Historical Note

Juan Parrondo discovered this paradox in 1996 while studying Brownian ratchets — theoretical devices that could extract useful work from random thermal motion. Richard Feynman had shown such devices couldn't work in thermal equilibrium, but Parrondo found that switching between two equilibrium states could produce directed motion.

The paradox has since been proven in physical experiments and found applications in biology, game theory, and even quantum computing.